Comparison of bulk sequencing, SMARTseq2 data and 10X single cell data
Load the packages used
library("tidyr")
library("stringr")
library("tibble")
library("pheatmap")
library("Seurat")Loading required package: ggplot2
Loading required package: cowplot
Attaching package: ‘cowplot’
The following object is masked from ‘package:ggplot2’:
ggsave
Loading required package: Matrix
Attaching package: ‘Matrix’
The following object is masked from ‘package:tidyr’:
expandlibrary("ggrepel")
library("dplyr")
Attaching package: ‘dplyr’
The following object is masked from ‘package:Biobase’:
combine
The following objects are masked from ‘package:BiocGenerics’:
combine, intersect, setdiff, union
The following objects are masked from ‘package:stats’:
filter, lag
The following objects are masked from ‘package:base’:
intersect, setdiff, setequal, unionlibrary("forcats")Load all the data
We have three different sequencing experiments and want to cross compare between
- Bulk sequencing of NSCs, Neuroblasts, Microglia and Endothelial Cells
- Single Cell RNA sequencing of Neural Stem cells with
- SMARTseq2 protocol: qNSC1 - aNSC2
- 10X protocol: qNSC1 - NB + OPC & OD
Define colors for young and old throughout the analysis
age_colors <- c(young = "yellowgreen" , old = "slateblue")
age_colors_bulk <- c(young = "yellowgreen" , old = "slateblue" , Mo7 = "slateblue1")Load the bulk sequencing TPM values
bulk_data <- read.csv(file = "../Bulk_sequencing/count_table/Gene_expression_matrix_population.csv" , row.names = 1)
bulk_data <- as.matrix(bulk_data)Log2 the TPM values
bulk_data <- log2(bulk_data+1)Also we load the annotation
bulk_annotation <- read.csv(file = "../Bulk_sequencing/annotation/annotation_bulk.csv" , row.names = 1)bulk_annotation$type <- factor(x = bulk_annotation$type , levels = c("NSC","Neuroblast","Microglia","Endothelial"))
bulk_annotation$age <- factor(x = bulk_annotation$age , levels = c("young","Mo7","old"))Which celltypes do we have?
bulk_celltypes <- unique(as.character(bulk_annotation$type))
bulk_celltypes[1] "NSC" "Neuroblast" "Endothelial" "Microglia" Load the SMARTseq2 sequencing TPM values
smartseq2_data <- read.csv(file = "../SmartSeq2/count_table/TPM_NSCs.csv") %>% dplyr::select(-X) %>% column_to_rownames(var = "ensembl_gene_id")
smartseq2_data <- as.matrix(smartseq2_data)Log2 the TPM values
smartseq2_data <- log2(smartseq2_data+1)Also we load the annotation
smartseq2_annotation <- read.csv(file = "../SmartSeq2/count_table/NSCs_annotation.csv" , row.names = 1)smartseq2_annotation$type <- fct_recode(.f = smartseq2_annotation$type , qNSC1 = "q1" , qNSC2 = "q2" , aNSC1 = "a1" , aNSC2 = "a2")
smartseq2_annotation$type <- factor(x = smartseq2_annotation$type , levels = c("qNSC1","qNSC2","aNSC1","aNSC2"))
smartseq2_annotation$age <- factor(x = smartseq2_annotation$age , levels = c("young","old"))Which celltypes do we have?
smartseq2_celltype_colors <- c( qNSC1 = "steelblue" , qNSC2 = "steelblue1" , aNSC1 = "tomato" , aNSC2 = "sienna1")Load the 10X sequencing normalized values
seurat_10X2 <- readRDS(file = "../10X/seurat_10X2_clustered_min_1500_nGene.RDS" )We can extract the normalized expression from the object
tenx_data <- expm1( seurat_10X2@data )
tenx_data <- as.matrix(tenx_data)Log2 the TPM values
tenx_data <- log2(tenx_data+1)Also we extract the annotation from the object
tenx_annotation <- FetchData(object = seurat_10X2 , vars.all = c("celltype","age") ) %>% rownames_to_column("cell") %>% dplyr::rename(type = celltype)tenx_annotation$type <- factor(x = tenx_annotation$type , levels = c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2","TAP","NB","OPC","OD"))
tenx_annotation$age <- factor(x = tenx_annotation$age , levels = c("young","old"))Which celltypes do we have?
tenx_celltype_colors <- c( qNSC1 = "steelblue" , qNSC2 = "steelblue1" , aNSC0 = "tomato" , aNSC1 = "sienna1", aNSC2 = "sienna3" , TAP = "green" , NB = "yellow" , OPC = "pink" , OD = "violet")Also we can load the genes.tsv file from the CellRanger Output which contains a mapping of ENSEMBL Gene IDs to the Gene names used in the 10X object
tenx_gene_2_ensembl_tbl <- read.table(file = "../10X/count_table/filtered_gene_bc_matrices_mex/mm10/genes.tsv") %>% dplyr::rename( ensembl_gene_id = V1 , gene_symbol = V2 )We can translate all the gene symbols to ENSEMBL Gene IDS
tenx_ensembl_gene_id <- unlist(lapply( X = rownames(tenx_data) , FUN = function(gene){
as.character( tenx_gene_2_ensembl_tbl[ which(tenx_gene_2_ensembl_tbl$gene_symbol == gene) , "ensembl_gene_id" ] )[1]
} ))Add the ENSEMBL Gene IDs as rownames
tenx_data_ensembl <- tenx_data
rownames(tenx_data_ensembl) <- tenx_ensembl_gene_idLoad differentially expressed gene lists
Bulk sequencing
NSCs
bulk_DE_genes_NSC <- read.csv(file = "../Bulk_sequencing/DESeq2_results_old_vs_young/GP_DESeq2_results_Old_vs_Yang.csv" , row.names = 1) %>% mutate(gene_symbol = str_to_title(gene_symbol) )bulk_DE_genes_NSC_Mo7_vs_young <- read.csv(file = "../Bulk_sequencing/DESeq2_results_old_vs_young/GP_DESeq2_results_M7_vs_Young.csv" , row.names = 1) %>% mutate(gene_symbol = str_to_title(gene_symbol) )Neuroblasts
bulk_DE_genes_NB <- read.csv(file = "../Bulk_sequencing/DESeq2_results_old_vs_young/PSA_DESeq2_results_Old_vs_Yang.csv" , row.names = 1) %>% mutate(gene_symbol = str_to_title(gene_symbol) )Microglia
bulk_DE_genes_Microglia <- read.csv(file = "../Bulk_sequencing/DESeq2_results_old_vs_young/CD11b_DESeq2_results_Old_vs_Yang.csv" , row.names = 1) %>% mutate(gene_symbol = str_to_title(gene_symbol) )Endothelial Cells
bulk_DE_genes_Endothelial <- read.csv(file = "../Bulk_sequencing/DESeq2_results_old_vs_young/CD31_DESeq2_results_Old_vs_Yang.csv" , row.names = 1) %>% mutate(gene_symbol = str_to_title(gene_symbol) )Make a list of the differentially expressed genes
bulk_DE_genes <- list( NSC = bulk_DE_genes_NSC , NB = bulk_DE_genes_NB , Microglia = bulk_DE_genes_Microglia , Endothelial = bulk_DE_genes_Endothelial )MA Plots
MAPlot <- function( x , main = "" , highlight_genes = c() , data = NULL , SMARTseq2 = FALSE , labeL_top_genes = FALSE , max_foldChange = 10 , min_baseMean = 0.1 , max_baseMean = 1000000 ){
require(dplyr)
require(ggplot2)
require(ggrepel)
if(!is.null(data)){
rowmean <- rowMeans(data)
rowmean_df <- data.frame(baseMean = as.numeric(rowmean) , gene_symbol = as.character( names(rowmean) ) )
if(SMARTseq2){
colnames(rowmean_df)[2] <- "ensembl_gene_id"
x <- left_join(x = x, y = rowmean_df , by = "ensembl_gene_id" )
}else{
x <- left_join(x = x, y = rowmean_df , by = "gene_symbol" )
}
}
x_sig <- filter( x , padj < 0.05 & abs(log2FoldChange) > 1 )
if( labeL_top_genes ){ x_label <- filter( x_sig , abs(log2FoldChange) > 4 ) }
x_highlight <- filter( x , gene_symbol %in% highlight_genes )
g <- ggplot(data = x , mapping = aes(x = baseMean , y = log2FoldChange )) + geom_point(color = "grey" ) + geom_point(data = x_sig , color = "red" ) + geom_point(data = x_highlight , color = "black", fill = "cyan" , size = 2 , pch = 21)
if( labeL_top_genes ){
g <- g + geom_text_repel(data = x_label , mapping = aes( x = baseMean , y = log2FoldChange , label = gene_symbol))
}
g <- g + scale_x_log10(limits = c(min_baseMean, max_baseMean ), expand = c(0, 0)) + ggtitle(main) + ylim(c(-max_foldChange,max_foldChange))
g
}Bulk
NSC
With Labels
MAPlot(x = bulk_DE_genes_NSC , main = "NSC (19 Month vs 2 Month)" , labeL_top_genes = TRUE )
MAPlot(x = bulk_DE_genes_NSC_Mo7_vs_young, main = "NSC (7 Month vs 2 Month)" , labeL_top_genes = TRUE)
Without Labels
MAPlot(x = bulk_DE_genes_NSC , main = "NSC (19 Month vs 2 Month)" )
MAPlot(x = bulk_DE_genes_NSC_Mo7_vs_young , main = "NSC (7 Month vs 2 Month)")
NB
MAPlot(x = bulk_DE_genes_NB , main = "NB")
MAPlot(x = bulk_DE_genes_NB , main = "NB" , labeL_top_genes = TRUE)
Microglia
MAPlot(x = bulk_DE_genes_Microglia , main = "Microglia")
MAPlot(x = bulk_DE_genes_Microglia , main = "Microglia" , labeL_top_genes = TRUE)
Endothelial cells
MAPlot(x = bulk_DE_genes_Endothelial , main = "Endothelial")
MAPlot(x = bulk_DE_genes_Endothelial , main = "Endothelial" , labeL_top_genes = TRUE)
Heatmaps of gene expression
HeatmapGenes <- function( x , genes = c() , topVar = NULL , main = "" , annotation , annotation_colors , cluster_cols = FALSE , cluster_rows = TRUE , show_rownames = FALSE , show_colnames = FALSE , arrange_by_age = FALSE , keep_non_expressed = FALSE ){
require("pheatmap")
require("viridisLite")
# The cells in the annotation are taken as the cells to be plotted
annotation <- remove_rownames(df = annotation)
if(! arrange_by_age){
annotation <- annotation %>% arrange( type , age)
}else{
annotation <- annotation %>% arrange( age , type)
}
if("cell" %in% colnames(annotation)){ names_cols <- annotation %>% pull("cell") %>% as.character() }
if("sample" %in% colnames(annotation)){ names_cols <- annotation %>% pull("sample") %>% as.character() }
# Then the annotation data.frame is prepared as needed for the pheatmap function with the cell names as rownames
if("cell" %in% colnames(annotation)){annotation <- column_to_rownames(df = annotation , var = "cell")}
if("sample" %in% colnames(annotation)){annotation <- column_to_rownames(df = annotation , var = "sample")}
# We set the order of columns of the annotation to be first age then type
annotation <- annotation[,c("age","type")]
# We check for the genes from the list that are available in our data
genes <- as.character(genes)
genes_avail <- genes[genes %in% rownames(x)]
if(!is.null(topVar)){
x <- x[,names_cols]
var_genes <- names(sort(apply(X = x, MARGIN = 1, FUN = var), decreasing = TRUE)[1:topVar])
var_genes <- var_genes[!is.na(var_genes)]
x_mat <- x[var_genes ,]
}else{
# Only take the values for genes that are from the gene list
x_mat <- x[genes_avail,names_cols]
}
# Dicard all the non expressed genes as they would create problems during the clustering process
if(!keep_non_expressed){
x <- x[apply(X = x, MARGIN = 1, FUN = var) > 0,]
}
if(keep_non_expressed & cluster_rows){warning("Row clustering cannot be used if unexpressed genes are kept!")}
pheatmap(
mat = x_mat ,
cluster_rows = cluster_rows ,
cluster_cols = cluster_cols ,
color = viridis(n = 100),
show_rownames = show_rownames,
show_colnames = show_colnames ,
clustering_method = "complete",
clustering_distance_rows = "euclidean" ,
border_color = NA,
scale = "none",
annotation_col = annotation,
main = main,
fontsize = 5,
annotation_colors = annotation_colors
)
}SMARTseq2 data
Top Variable Genes in subpopulations
qNSC1: Top 5000 variable genes
HeatmapGenes(x = smartseq2_data , topVar = 5000 , main = "SMARTseq2: qNSC1 young and old Top variable genes" , annotation = filter(smartseq2_annotation , type %in% c("qNSC1")) , annotation_colors = list(age = age_colors , type = smartseq2_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)Loading required package: viridisLite
qNSC2: Top 5000 variable genes
HeatmapGenes(x = smartseq2_data , topVar = 5000 , main = "SMARTseq2: qNSC2 young and old Top variable genes" , annotation = filter(smartseq2_annotation , type %in% c("qNSC2")) , annotation_colors = list(age = age_colors , type = smartseq2_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)
aNSC1: Top 5000 variable genes
HeatmapGenes(x = smartseq2_data , topVar = 5000 , main = "SMARTseq2: aNSC1 young and old Top variable genes" , annotation = filter(smartseq2_annotation , type %in% c("aNSC1")) , annotation_colors = list(age = age_colors , type = smartseq2_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)
aNSC2: Top 5000 variable genes
HeatmapGenes(x = smartseq2_data , topVar = 5000 , main = "SMARTseq2: aNSC2 young and old Top variable genes" , annotation = filter(smartseq2_annotation , type %in% c("aNSC2")) , annotation_colors = list(age = age_colors , type = smartseq2_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)
10X data
Top Variable Genes in subpopulations
qNSC1: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: qNSC1 young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("qNSC1")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)qNSC2: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: qNSC2 young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("qNSC2")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE )aNSC0: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: aNSC0 young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("aNSC0")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)aNSC1: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: aNSC1 young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("aNSC1")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)aNSC2: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: aNSC2 young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("aNSC2")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)TAP: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: TAP young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("TAP")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)NB: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: NB young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("NB")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)OPC: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: OPC young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("OPC")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)OD: Top 5000 variable genes
# HeatmapGenes(x = tenx_data_ensembl , topVar = 5000 , main = "10X: OD young and old Top variable genes" , annotation = filter(tenx_annotation , type %in% c("OD")) , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE , cluster_rows = TRUE)Compare pseudo-bulk Single Cell RNA-seq with bulk samples - mean TPM values
Pseudobulk_bulk_compare <- function( pseudobulk_data = NULL , pseudobulk_celltype = "x" , bulk_celltype = "y" , bulk_data = NULL , annotation_pseudobulk = NULL , annotation_bulk = NULL , celltype_bulk = NULL , main = "" ){
require("ggplot2")
require("dplyr")
require("tibble")
if( is.null(pseudobulk_data) | is.null(bulk_data) ){ stop("You need to supply all data !") }
if( is.null(annotation_pseudobulk) | is.null(annotation_pseudobulk) ){stop("You need to supply the annotation !")}
## Strip Log2 transform of the data
pseudobulk_data <- 2^(pseudobulk_data)-1
bulk_data <- 2^(bulk_data)-1
pseudobulk_data <- pseudobulk_data[!is.na(rownames(pseudobulk_data)),]
bulk_data <- bulk_data[!is.na(rownames(bulk_data)),]
pseudobulk_data <- pseudobulk_data[,annotation_pseudobulk$cell]
bulk_data <- bulk_data[,annotation_bulk$sample]
pseudobulk_data_comb <- data.frame( expr_x = log2(rowMeans(pseudobulk_data)+1) , ensembl_gene_id = rownames(pseudobulk_data) )
bulk_data_comb <- data.frame( expr_y = log2(rowMeans(bulk_data)+1) , ensembl_gene_id = rownames(bulk_data) )
x <- full_join(x = pseudobulk_data_comb , y = bulk_data_comb , by = "ensembl_gene_id" )
g <- ggplot(data = x , mapping = aes(x = expr_x , y = expr_y)) + geom_point(alpha = 0.5) + coord_equal() + labs(x = pseudobulk_celltype , y = bulk_celltype ) + ggtitle(main)
g
}SMARTseq2
Pseudobulk_bulk_compare(pseudobulk_data = smartseq2_data , bulk_data = bulk_data , annotation_pseudobulk = filter(smartseq2_annotation, type %in% c("qNSC1","qNSC2","aNSC1","aNSC2"), age == "young") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "young"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(TPM+1))" , bulk_celltype = " bulk samples: NSC (log2(TPM+1))" , main = "Young NSC: pseudo-bulk vs bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk_bulk_compare(pseudobulk_data = smartseq2_data , bulk_data = bulk_data , annotation_pseudobulk = filter(smartseq2_annotation, type %in% c("qNSC1","qNSC2","aNSC1","aNSC2"), age == "old") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "young"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(TPM+1))" , bulk_celltype = " bulk samples: NSC (log2(TPM+1))" , main = "NSC: old pseudo-bulk (SMARTseq2) vs young bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk_bulk_compare(pseudobulk_data = smartseq2_data , bulk_data = bulk_data , annotation_pseudobulk = filter(smartseq2_annotation, type %in% c("qNSC1","qNSC2","aNSC1","aNSC2"), age == "young") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "old"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(TPM+1))" , bulk_celltype = " bulk samples: NSC (log2(TPM+1))" , main = "NSC: young pseudo-bulk (SMARTseq2) vs old bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk_bulk_compare(pseudobulk_data = smartseq2_data , bulk_data = bulk_data , annotation_pseudobulk = filter(smartseq2_annotation, type %in% c("qNSC1","qNSC2","aNSC1","aNSC2"), age == "old") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "old"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(TPM+1))" , bulk_celltype = " bulk samples: NSC (log2(TPM+1))" , main = "Old NSC: pseudo-bulk (SMARTseq2) vs bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
10X
Pseudobulk_bulk_compare(pseudobulk_data = tenx_data_ensembl , bulk_data = bulk_data , annotation_pseudobulk = filter(tenx_annotation, type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2"), age == "old") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "old"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(expression+1))" , bulk_celltype = " bulk samples: NSC (log2(TPM+1))" , main = "Old NSC: pseudo-bulk (10X) vs bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk_bulk_compare(pseudobulk_data = tenx_data_ensembl , bulk_data = bulk_data , annotation_pseudobulk = filter(tenx_annotation, type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2"), age == "old") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "young"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(expression+1))" , bulk_celltype = " bulk samples: NSC (log2(TPM+1))" , main = "NSC: Old pseudo-bulk (10X) vs young bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk_bulk_compare(pseudobulk_data = tenx_data_ensembl , bulk_data = bulk_data , annotation_pseudobulk = filter(tenx_annotation, type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2"), age == "young") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "old"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(TPM+1))" , bulk_celltype = " bulk samples: NSC (log2(expression+1))" , main = "NSC: Young pseudo-bulk (10X) vs old bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk_bulk_compare(pseudobulk_data = tenx_data_ensembl , bulk_data = bulk_data , annotation_pseudobulk = filter(tenx_annotation, type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2"), age == "young") , annotation_bulk = filter(bulk_annotation, type %in% c("NSC") , age == "young"), pseudobulk_celltype = "pseudo-bulk samples: NSC (log2(TPM+1))" , bulk_celltype = " bulk samples: NSC (log2(expression+1))" , main = "Young NSC: pseudo-bulk (10X) vs bulk" )Column `ensembl_gene_id` joining factors with different levels, coercing to character vector
Pseudobulk MA Plot
Pseudobulk_bulk_MAPlot <- function( pseudobulk_data = NULL , annotation_pseudobulk = NULL , main = "" , labeL_top_genes = FALSE , highlight_genes = c() , max_foldChange = 10 , max_baseMean = 1000000 ){
require("ggplot2")
require("dplyr")
require("tibble")
## Strip Log2 transform of the data
pseudobulk_data <- 2^(pseudobulk_data) - 1
bulk_data <- 2^(bulk_data) - 1
if( is.null(pseudobulk_data) ){ stop("You need to supply all data !") }
if( is.null(annotation_pseudobulk) ){stop("You need to supply the annotation !")}
annotation_pseudobulk_old <- filter(annotation_pseudobulk , age == "old")
annotation_pseudobulk_young <- filter(annotation_pseudobulk , age == "young")
pseudobulk_data <- pseudobulk_data[!is.na(rownames(pseudobulk_data)),]
pseudobulk_data_old <- pseudobulk_data[,as.character(annotation_pseudobulk_old$cell)]
pseudobulk_data_young <- pseudobulk_data[,as.character(annotation_pseudobulk_young$cell)]
pseudobulk_data_complete <- pseudobulk_data[, c( as.character(annotation_pseudobulk_old$cell),as.character(annotation_pseudobulk_young$cell)) ]
if( ! all(rownames(pseudobulk_data_old) == rownames(pseudobulk_data_young))){stop("Rownames are not correct: Pseudobulk")}
pseudobulk_data_comb <- data.frame( log2FoldChange = log2(((rowSums(pseudobulk_data_old) + 1) / (rowSums(pseudobulk_data_young) + 1) ) ) , expr_young_pseudobulk = rowSums(pseudobulk_data_young) , baseMean = rowMeans(pseudobulk_data_complete), expr_old_pseudobulk = rowSums(pseudobulk_data_old) , gene = rownames(pseudobulk_data_old) )
pseudobulk_data_comb$log2FoldChange[is.nan(pseudobulk_data_comb$log2FoldChange)] <- 0
x_highlight <- filter(pseudobulk_data_comb ,gene %in% highlight_genes )
g <- ggplot(data = pseudobulk_data_comb , mapping = aes(x = baseMean , y = log2FoldChange)) + labs(x = "baseMean" , y = "Log2FoldChange" ) + ggtitle(main) + geom_vline(xintercept = 0) + geom_hline(yintercept = 0) + geom_point(color = "grey" ) + geom_point(data = x_highlight , color = "black", fill = "cyan" , size = 2 , pch = 21)
if( labeL_top_genes ){
g <- g + geom_text_repel(data = x_label , mapping = aes( x = baseMean , y = log2FoldChange , label = gene))
}
g <- g + scale_x_log10(limits = c(0.00001, max_baseMean ), expand = c(0, 0)) + ggtitle(main) + ylim(c(-max_foldChange,max_foldChange))
g
}Pseudobulk_bulk_MAPlot(pseudobulk_data = smartseq2_data , annotation_pseudobulk = smartseq2_annotation)
Pseudobulk_bulk_MAPlot(pseudobulk_data = tenx_data , annotation_pseudobulk = dplyr::filter(tenx_annotation, type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2")) )
Plot individual genes in the single cell datasets
Beeswarm plot of Il33
SMARTseq2
IL33_smartseq2_data <- smartseq2_data["ENSMUSG00000024810",]IL33_smartseq2_df <- data.frame( expression = IL33_smartseq2_data , row.names = names(IL33_smartseq2_data) ) %>% rownames_to_column("cell") %>% left_join( smartseq2_annotation , by = "cell")Column `cell` joining character vector and factor, coercing into character vectoraddmargins( table( IL33_smartseq2_df$age , IL33_smartseq2_df$expression > 2.5 ) )
FALSE TRUE Sum
young 73 16 89
old 71 62 133
Sum 144 78 222library(ggbeeswarm)
ggplot(data = IL33_smartseq2_df , mapping = aes(x = age , y = expression , color = type )) + geom_quasirandom( size = 1.5 ,method = "smiley" ) + xlab("Age") + ylab("log2(TPM+1)") + ggtitle("Expression of Il33 in the SMARTseq2 data") + scale_x_discrete(labels = c(young = "2 month" , old = "22 month") ) + scale_color_manual( name = "Activation state" , values = smartseq2_celltype_colors ) + guides(color = guide_legend(override.aes=list(size=1.5)))
10X
IL33_10X_data <- tenx_data_ensembl["ENSMUSG00000024810",]tenx_annotation_nscs <- tenx_annotation %>% filter( type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2"))
IL33_10X_df <- data.frame( expression = IL33_10X_data , row.names = names(IL33_10X_data) ) %>% rownames_to_column("cell") %>% filter(cell %in% as.character( tenx_annotation_nscs$cell ) ) %>% left_join( tenx_annotation_nscs , by = "cell")addmargins( table( IL33_10X_df$age , IL33_10X_df$expression > 0.5 ) )
FALSE TRUE Sum
young 1162 63 1225
old 836 185 1021
Sum 1998 248 2246library(ggbeeswarm)
ggplot(data = IL33_10X_df , mapping = aes(x = age , y = expression , color = type )) + xlab("Age") + ylab("log2(expression+1)") + ggtitle("Expression of Il33 in the 10X data") + scale_x_discrete(labels = c(young = "2 month" , old = "22 month") ) + scale_color_manual( name = "Activation state" , values = tenx_celltype_colors ) + guides(color = guide_legend(override.aes=list(size=1.5))) + geom_quasirandom( size = 0.5 ,method = "smiley" ) 
Search for genes from Gene Lists in all datasets
InnateDB genes - manually curated from the bulk seq data (from sig. genes with log2FoldChange > 1)
DE_genes_NSCs_bulk_up_InnateDB <- read.csv("GeneLists/Manually_curated_DE_genes_from_bulk/InnateDB_Gene_list/InnateDB_NSCs.csv" , stringsAsFactors = FALSE , header = TRUE)
library(biomaRt)
mart <- useMart(biomart = "ensembl", dataset = "mmusculus_gene_ensembl")
genes_symbol_ensembl <- getBM(attributes = c("mgi_symbol","ensembl_gene_id" ), filters = "mgi_symbol" , values = DE_genes_NSCs_bulk_up_InnateDB$Genes , mart = mart )
genes_symbol_ensembl <- dplyr::rename(genes_symbol_ensembl , MGI.symbol = mgi_symbol , gene = ensembl_gene_id )
genes_symbol_ensembl <- dplyr::filter( genes_symbol_ensembl , gene %in% as.character(bulk_DE_genes_NSC %>% filter(padj < 0.05 , log2FoldChange > 1) %>% pull("ensembl_gene_id") ) )
DE_genes_NSCs_bulk_up_InnateDB <- genes_symbol_ensemblgene_set <- DE_genes_NSCs_bulk_up_InnateDBplot_title <- "NSC bulk seq (top DE genes between old and young) annotated in InnateDB"Bulk
Bulk sequencing: Heatmap
HeatmapGenes(x = bulk_data , genes = unique(gene_set$gene) , main = paste("Bulk:", plot_title) , annotation = bulk_annotation , show_colnames = FALSE , annotation_colors = list(age = age_colors_bulk ) )
Bulk sequencing: MA-Plot
NSC
MAPlot(x = bulk_DE_genes_NSC , main = "NSC (19 month vs 2 month)" , highlight_genes = unique(gene_set$MGI.symbol) ) 
MAPlot(x = bulk_DE_genes_NSC_Mo7_vs_young , main = "NSC (7 Month vs 2 Month)" , highlight_genes = unique(gene_set$MGI.symbol) ) 
NB
MAPlot(x = bulk_DE_genes_NB , main = "NB" , highlight_genes = unique(gene_set$MGI.symbol) ) 
Microglia
MAPlot(x = bulk_DE_genes_Microglia , main = "Microglia" , highlight_genes = unique(gene_set$MGI.symbol) ) 
Endothelial cells
MAPlot(x = bulk_DE_genes_Endothelial , main = "Endothelial" , highlight_genes = unique(gene_set$MGI.symbol) ) 
Pseudobulk SMARTseq2
Pseudobulk_bulk_MAPlot(pseudobulk_data = smartseq2_data , annotation_pseudobulk = smartseq2_annotation , highlight_genes = unique(gene_set$gene) , main = "Pseudobulk: SMARTseq2")
Pseudobulk 10X
Pseudobulk_bulk_MAPlot(pseudobulk_data = tenx_data , annotation_pseudobulk = dplyr::filter(tenx_annotation, type %in% c("qNSC1","qNSC2","aNSC0","aNSC1","aNSC2")) , highlight_genes = unique(gene_set$MGI.symbol) , main = "Pseudobulk: 10X")
SMARTseq2
SMARTseq2 sequencing: Heatmap
Arranged by celltype
HeatmapGenes(x = smartseq2_data , genes = unique(gene_set$gene) , main = paste("SMARTseq2:", plot_title) , annotation = smartseq2_annotation , annotation_colors = list(age = age_colors , type = smartseq2_celltype_colors ) )
Columns clustered
test <- HeatmapGenes(x = smartseq2_data , genes = unique(gene_set$gene) , main = paste("SMARTseq2:", plot_title) , annotation = smartseq2_annotation %>% dplyr::filter(type %in% c("qNSC1","qNSC2")) , annotation_colors = list(age = age_colors , type = smartseq2_celltype_colors ) , cluster_cols = TRUE )
10X
10X sequencing: Heatmap
HeatmapGenes(x = tenx_data_ensembl , genes = unique(gene_set$gene) , main = paste("10X:", plot_title) , annotation = tenx_annotation , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) )
HeatmapGenes(x = tenx_data_ensembl , genes = unique(gene_set$gene) , main = paste("10X:", plot_title) , annotation = tenx_annotation , annotation_colors = list(age = age_colors , type = tenx_celltype_colors ) , cluster_cols = TRUE )
InnateDB genes - manually curated from the bulk seq data (from sig. genes with log2FoldChange > 1) - from Endothelial, NB and Microglia
Endothelial
DE_genes_Endothelial_bulk_up_InnateDB <- read.csv("GeneLists/Manually_curated_DE_genes_from_bulk/InnateDB_Gene_list/InnateDB_Endos.csv" , stringsAsFactors = FALSE , header = TRUE)
library(biomaRt)
mart <- useMart(biomart = "ensembl", dataset = "mmusculus_gene_ensembl")
genes_symbol_ensembl <- getBM(attributes = c("mgi_symbol","ensembl_gene_id" ), filters = "mgi_symbol" , values = DE_genes_Endothelial_bulk_up_InnateDB$Genes , mart = mart )
genes_symbol_ensembl <- dplyr::rename(genes_symbol_ensembl , MGI.symbol = mgi_symbol , gene = ensembl_gene_id )
genes_symbol_ensembl <- dplyr::filter( genes_symbol_ensembl , gene %in% as.character(bulk_DE_genes_Endothelial %>% filter(padj < 0.05 , log2FoldChange > 1) %>% pull("ensembl_gene_id") ) )
DE_genes_Endothelial_bulk_up_InnateDB <- genes_symbol_ensemblMAPlot(x = bulk_DE_genes_Endothelial , main = "Endothelial" , highlight_genes = unique(DE_genes_Endothelial_bulk_up_InnateDB$MGI.symbol) ) 
Microglia
DE_genes_Microglia_bulk_up_InnateDB <- read.csv("GeneLists/Manually_curated_DE_genes_from_bulk/InnateDB_Gene_list/InnateDB_Microglia.csv" , stringsAsFactors = FALSE , header = TRUE)
library(biomaRt)
mart <- useMart(biomart = "ensembl", dataset = "mmusculus_gene_ensembl")
genes_symbol_ensembl <- getBM(attributes = c("mgi_symbol","ensembl_gene_id" ), filters = "mgi_symbol" , values = DE_genes_Microglia_bulk_up_InnateDB$Genes , mart = mart )
genes_symbol_ensembl <- dplyr::rename(genes_symbol_ensembl , MGI.symbol = mgi_symbol , gene = ensembl_gene_id )
genes_symbol_ensembl <- dplyr::filter( genes_symbol_ensembl , gene %in% as.character(bulk_DE_genes_Microglia %>% dplyr::filter(padj < 0.05 , log2FoldChange > 1) %>% pull("ensembl_gene_id") ) )
DE_genes_Microglia_bulk_up_InnateDB <- genes_symbol_ensemblMAPlot(x = bulk_DE_genes_Microglia , main = "Microglia" , highlight_genes = unique(DE_genes_Microglia_bulk_up_InnateDB$MGI.symbol) ) 
NB
DE_genes_NB_bulk_up_InnateDB <- read.csv("GeneLists/Manually_curated_DE_genes_from_bulk/InnateDB_Gene_list/InnateDB_Neuroblasts.csv" , stringsAsFactors = FALSE , header = TRUE)
library(biomaRt)
mart <- useMart(biomart = "ensembl", dataset = "mmusculus_gene_ensembl")
genes_symbol_ensembl <- getBM(attributes = c("mgi_symbol","ensembl_gene_id" ), filters = "mgi_symbol" , values = DE_genes_NB_bulk_up_InnateDB$Genes , mart = mart )
genes_symbol_ensembl <- dplyr::rename(genes_symbol_ensembl , MGI.symbol = mgi_symbol , gene = ensembl_gene_id )
genes_symbol_ensembl <- dplyr::filter( genes_symbol_ensembl , gene %in% as.character(bulk_DE_genes_NB %>% dplyr::filter(padj < 0.05 , log2FoldChange > 1) %>% pull("ensembl_gene_id") ) )
DE_genes_NB_bulk_up_InnateDB <- genes_symbol_ensemblMAPlot(x = bulk_DE_genes_NB , main = "Neuroblast" , highlight_genes = unique(DE_genes_NB_bulk_up_InnateDB$MGI.symbol) ) 
Load raw counts
We can also compare the raw counts for specific genes that we recover with the different sequencing methods
Bulk sequencing data
data_bulk <- read.csv(file = "raw_counts/compatible_rownames_only_NSCs/Bulk_seq_All.genes.counts_compat.csv" , row.names = 1)
anno_bulk <- read.csv(file = "raw_counts/annotation_bulk.csv" , row.names = 1)Check if all cells from the annotation are present in the data as columns and the other way around
all(colnames(data_bulk) == anno_bulk$name)[1] TRUESMARTseq2 data
data_SMARTseq2_NSCs <- read.csv(file = "raw_counts/compatible_rownames_only_NSCs/SMARTseq2_All.genes.counts_compat.csv" , row.names = 1)
anno_SMARTseq2_NSCs <- read.csv(file = "raw_counts/annotation_NSCs_SMARTseq2.csv" , row.names = 1)Check if all cells from the annotation are present in the data as columns and the other way around
all(colnames(data_SMARTseq2_NSCs) == anno_SMARTseq2_NSCs$name)[1] TRUE10X data
data_10X_NSCs <- read.csv(file = "raw_counts/compatible_rownames_only_NSCs/10X_All.genes.counts_compat.csv" , row.names = 1)
anno_10X_NSCs <- read.csv(file = "raw_counts/annotation_10x_NSCs.csv" , row.names = 1)colnames(data_10X_NSCs) <- sub(x = colnames(data_10X_NSCs) , pattern = "\\." , replacement = "-" )Check if all cells from the annotation are present in the data as columns and the other way around
all(colnames(data_10X_NSCs) == anno_10X_NSCs$name)[1] TRUECheck raw counts for highly upregulated genes in bulk NSC old to young comparison
10X
How many counts do genes have in the Single cell data sets, that are significantly upregulated in the bulk NSCs
cells_NSCs_10X <- as.character(anno_10X_NSCs$name)Iigp1 - ENSMUSG00000054072
Iigp1_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000054072",cells_NSCs_10X]) , gene = "Iigp1" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Iigp1_df$age , Iigp1_df$geneReads > 0 ) )
FALSE TRUE Sum
old 996 25 1021
young 1221 4 1225
Sum 2217 29 2246gg_Iigp1 <- ggplot(data = Iigp1_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Iigp1
Ddx60 - ENSMUSG00000037921
Ddx60_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000037921",cells_NSCs_10X]) , gene = "Ddx60" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ddx60_df$age , Ddx60_df$geneReads > 0 ) )
FALSE TRUE Sum
old 1002 19 1021
young 1224 1 1225
Sum 2226 20 2246gg_Ddx60 <- ggplot(data = Ddx60_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )Ifit1 - ENSMUSG00000078920
Ifit1_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000078920",cells_NSCs_10X]) , gene = "Ifit1" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ifit1_df$age , Ifit1_df$geneReads > 0 ) )
FALSE TRUE Sum
old 1017 4 1021
young 1223 2 1225
Sum 2240 6 2246gg_Ifit1 <- ggplot(data = Ifit1_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ifit1
Rsad2 - ENSMUSG00000020641
Rsad2_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000020641",cells_NSCs_10X]) , gene = "Rsad2" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Rsad2_df$age , Rsad2_df$geneReads > 0 ) )
FALSE TRUE Sum
old 995 26 1021
young 1213 12 1225
Sum 2208 38 2246gg_Rsad2 <- ggplot(data = Rsad2_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Rsad2
C3 - ENSMUSG00000024164
C3_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000024164",cells_NSCs_10X]) , gene = "C3" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( C3_df$age , C3_df$geneReads > 0 ) )
FALSE Sum
old 1021 1021
young 1225 1225
Sum 2246 2246gg_C3 <- ggplot(data = C3_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_C3
Ifi44 - ENSMUSG00000028037
Ifi44_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000028037",cells_NSCs_10X]) , gene = "Ifi44" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ifi44_df$age , Ifi44_df$geneReads > 0 ) )
FALSE TRUE Sum
old 1019 2 1021
young 1223 2 1225
Sum 2242 4 2246gg_Ifi44 <- ggplot(data = Ifi44_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ifi44
Cxcl10 - ENSMUSG00000034855
Cxcl10_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000034855",cells_NSCs_10X]) , gene = "Cxcl10" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Cxcl10_df$age , Cxcl10_df$geneReads > 0 ) )
FALSE TRUE Sum
old 954 67 1021
young 1154 71 1225
Sum 2108 138 2246gg_Cxcl10 <- ggplot(data = Cxcl10_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Cxcl10
Il33 - ENSMUSG00000024810
Il33_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000024810",cells_NSCs_10X]) , gene = "Il33" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Il33_df$age , Il33_df$geneReads > 0 ) )
FALSE TRUE Sum
old 836 185 1021
young 1162 63 1225
Sum 1998 248 2246gg_Il33 <- ggplot(data = Il33_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Il33
Ifi47 - ENSMUSG00000078920
Ifi47_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000078920",cells_NSCs_10X]) , gene = "Ifi47" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ifi47_df$age , Ifi47_df$geneReads > 0 ) )
FALSE TRUE Sum
old 1017 4 1021
young 1223 2 1225
Sum 2240 6 2246gg_Ifi47 <- ggplot(data = Ifi47_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ifi47
Gm4951 - ENSMUSG00000073555
Gm4951_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000073555",cells_NSCs_10X]) , gene = "Gm4951" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Gm4951_df$age , Gm4951_df$geneReads > 0 ) )
FALSE TRUE Sum
old 993 28 1021
young 1223 2 1225
Sum 2216 30 2246gg_Gm4951 <- ggplot(data = Gm4951_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Gm4951
Slc1a3 - ENSMUSG00000005360
Slc1a3_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000005360",cells_NSCs_10X]) , gene = "Slc1a3" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Slc1a3_df$age , Slc1a3_df$geneReads > 0 ) )
FALSE TRUE Sum
old 26 995 1021
young 48 1177 1225
Sum 74 2172 2246gg_Slc1a3 <- ggplot(data = Slc1a3_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Slc1a3
Sox2 - ENSMUSG00000074637
Sox2_df <- data.frame( totalReads = colSums(data_10X_NSCs) , geneReads = as.numeric(data_10X_NSCs["ENSMUSG00000074637",cells_NSCs_10X]) , gene = "Sox2" ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Sox2_df$age , Sox2_df$geneReads > 0 ) )
FALSE TRUE Sum
old 209 812 1021
young 202 1023 1225
Sum 411 1835 2246gg_Sox2 <- ggplot(data = Sox2_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Sox2
interferon_genes_bulk_df <- bind_rows( Ifit1_df , C3_df , Rsad2_df , Iigp1_df , Ifi47_df , Il33_df , Cxcl10_df , Sox2_df , Slc1a3_df )Unequal factor levels: coercing to characterbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorCombine all genes
Scatterplot
gg_interferon_genes <- ggplot(data = interferon_genes_bulk_df , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") ) + facet_grid( gene ~ . )gg_interferon_genes
Barplot
Show the same data as stacked barplots
interferon_genes_bulk_df_bars <- interferon_genes_bulk_df %>% mutate( interval = base::cut( geneReads , breaks = c(-0.1, 0.5,5.5,50.5,Inf) , labels = c("0","1-5","5-50","50-") ) ) %>% group_by(interval,gene,age) %>% summarise( counts = n() ) %>% ungroup() %>% mutate( interval = fct_rev( interval ) , age = fct_recode( .f = age , "22 month" = "old", "2 month" = "young" ) %>% fct_rev() ) %>% group_by(gene,age) %>% mutate( Percent = 100*counts/sum(counts) )ggplot( data = interferon_genes_bulk_df_bars , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ gene ) + viridis::scale_fill_viridis( discrete = TRUE , direction = -1 ) + ggtitle("Read counts: 10X")
Bulk NSC genes in InnateDB
Which genes are the ones with high fold changes in the bulk and are annotated in InnateDB
innateDB_genes_NSCs_bulk_up_table <- structure(list(ensembl_gene_id = structure(c(24L, 25L, 9L, 40L,
34L, 15L, 17L, 31L, 22L, 30L, 26L, 20L, 19L, 33L, 27L, 4L, 16L,
8L, 7L, 29L, 35L, 1L, 12L, 2L, 32L, 13L, 36L, 10L, 3L, 28L, 21L,
39L, 23L, 11L, 14L, 6L, 5L, 18L, 38L, 37L), .Label = c("ENSMUSG00000000386",
"ENSMUSG00000002325", "ENSMUSG00000005413", "ENSMUSG00000014361",
"ENSMUSG00000015766", "ENSMUSG00000017716", "ENSMUSG00000018899",
"ENSMUSG00000020122", "ENSMUSG00000020641", "ENSMUSG00000021508",
"ENSMUSG00000022306", "ENSMUSG00000023951", "ENSMUSG00000024066",
"ENSMUSG00000024079", "ENSMUSG00000024164", "ENSMUSG00000025492",
"ENSMUSG00000025498", "ENSMUSG00000026104", "ENSMUSG00000026896",
"ENSMUSG00000028270", "ENSMUSG00000029167", "ENSMUSG00000029826",
"ENSMUSG00000030341", "ENSMUSG00000034459", "ENSMUSG00000034855",
"ENSMUSG00000035692", "ENSMUSG00000036905", "ENSMUSG00000036913",
"ENSMUSG00000037860", "ENSMUSG00000038418", "ENSMUSG00000038642",
"ENSMUSG00000040747", "ENSMUSG00000046688", "ENSMUSG00000046718",
"ENSMUSG00000053113", "ENSMUSG00000054717", "ENSMUSG00000060441",
"ENSMUSG00000064215", "ENSMUSG00000074151", "ENSMUSG00000074896"
), class = "factor"), baseMean = c(1424.95976205735, 146.693384980886,
254.747257624993, 1378.74240836277, 232.679222864689, 53.3856147003376,
127.295243864467, 100.858827865452, 1254.48226869821, 8484.55808366433,
133.391584778487, 449.698563109268, 435.970338794981, 487.652611748752,
38.1324871848669, 2270.44585158928, 392.455793416614, 8123.4459446361,
495.513923324155, 147.634742373719, 285.437837120322, 20.0485104704833,
243.691996897216, 513.924995107202, 17.4190622356759, 18.1436595206331,
15890.0096695135, 1952.66573537053, 149.127545170108, 284.046380761723,
1257.30518127357, 52.8851930781209, 761.966603230871, 315.283958297394,
633.932254776546, 1091.87572537937, 450.61462668542, 1757.25022917456,
2425.56654256373, 64.6090952815893), log2FoldChange = c(4.55242859938374,
6.95678411131362, 5.03641775134329, 3.37966257532159, 3.73428610778296,
5.53141829606081, 3.86121962444797, 3.75531459127234, 2.42356052609584,
2.16374238670928, 3.1971506351276, 2.44063114014779, 2.44677858670807,
2.0893762261114, 3.67190586863129, 1.88755481055364, 2.10038800171168,
1.69845072572928, 1.91183744819457, 2.41403420088619, 2.0284929496982,
3.7014584495119, 1.94105240967367, 1.98333192338338, 3.4448344969081,
3.44363682239329, 1.57281037684989, 1.63883482734508, 2.11717369111306,
1.8668146256893, 1.43386187193389, 2.76038968668686, 1.7580616439951,
1.52810896028218, 1.55394451633379, 1.28056799567681, 1.42242513375296,
1.26735536508083, 1.18570808766668, 2.12810319958397), lfcSE = c(0.519384243203232,
0.81388488255093, 0.641951444263734, 0.48612175786269, 0.578514935508735,
0.878589568240684, 0.670578673894941, 0.684535958284751, 0.448804291965827,
0.41379046101385, 0.645816741770005, 0.504575112575078, 0.531966043635904,
0.460946382208802, 0.845572909706191, 0.436183111157877, 0.505011402058278,
0.415807705611759, 0.488183380722589, 0.618808403224005, 0.520676205372646,
0.965738215947367, 0.521153307010798, 0.553367834874403, 0.973891191367301,
0.97448322263972, 0.448895240692065, 0.468806743609123, 0.606652875776758,
0.541455272040843, 0.42118712809647, 0.818497304894856, 0.535426580260121,
0.488159657804439, 0.500868295142623, 0.422276770181993, 0.470630833485906,
0.427622347529472, 0.406142195700318, 0.7298433146452), stat = c(8.76504949651004,
8.54762664900375, 7.84548083246336, 6.95229645795079, 6.45495194432465,
6.29579327596285, 5.75804118854651, 5.48592743130992, 5.40003865711778,
5.2290774934922, 4.95055397041132, 4.83700261729543, 4.59950144558989,
4.53279666953747, 4.34250651420131, 4.32743671698567, 4.1590902564796,
4.08470238239196, 3.91622804808462, 3.90110119434225, 3.89588179518673,
3.83277619999832, 3.7245324620639, 3.58411132413133, 3.53718621489091,
3.53380821997635, 3.50373591492099, 3.49575779292008, 3.48992607741657,
3.44777255312879, 3.40433450189738, 3.37250919481214, 3.28347846149326,
3.13034667214217, 3.10250125912102, 3.03253241973247, 3.02237981990518,
2.96372575568792, 2.91944077768661, 2.91583571005033), pvalue = c(1.86694618820708e-18,
1.25645385522089e-17, 4.31297720037574e-15, 3.59387157612209e-12,
1.08253248479473e-10, 3.05831732678035e-10, 8.50955946071268e-09,
4.11305732780317e-08, 6.66265384184588e-08, 1.70357951680874e-07,
7.40025308679816e-07, 1.31811693167268e-06, 4.23503254113743e-06,
5.82078401153992e-06, 1.40866307399552e-05, 1.50854667946568e-05,
3.19517635376871e-05, 4.41333464318532e-05, 8.99451714694149e-05,
9.57561097095347e-05, 9.78421013854258e-05, 0.000126705214745366,
0.000195677532925451, 0.000338227814291896, 0.000404414420179005,
0.000409618180066453, 0.0004587800815838, 0.000472717577863198,
0.000483154175333289, 0.000565229800758029, 0.000663254779320521,
0.000744865993665538, 0.00102534479496883, 0.00174600119776919,
0.00191892731640661, 0.00242511046065225, 0.00250795646620197,
0.00303939026229686, 0.00350660015484411, 0.00354737241964209
), padj = c(2.76849450249227e-15, 1.43322724762081e-14, 3.55317438357621e-12,
1.15855481744162e-09, 2.43225366924561e-08, 5.96734047879288e-08,
1.10691453721849e-06, 4.09345819557002e-06, 6.02442035492272e-06,
1.40346559193094e-05, 5.05706695963733e-05, 7.72583240307279e-05,
0.00020864218455989, 0.000272291501915222, 0.000555143310344139,
0.00058560834318839, 0.00109173433525429, 0.00143520481192533,
0.00261017015209384, 0.00274124971212874, 0.00278483785306042,
0.00341326958812737, 0.00486047258919852, 0.007542225952082,
0.00869139338671661, 0.00873989639166249, 0.00959555688266032,
0.00981782767805792, 0.00993716125661212, 0.0113267469127579,
0.0128318926960938, 0.0140529488804914, 0.0180028284113852, 0.0273477266839752,
0.0294878478497343, 0.034989937779684, 0.0357257314479433, 0.0415793911120156,
0.0457740965635417, 0.0462249434190444), gene_symbol = c("Ifit1",
"Cxcl10", "Rsad2", "Ifit3", "Bst2", "C3", "Irf7", "Ctss", "Zc3hav1",
"Egr1", "Isg15", "Gbp2", "Ifih1", "Tifa", "C1qb", "Mertk", "Ifitm3",
"Egfr", "Irf1", "Aim2", "Socs3", "Mx1", "Vegfa", "Irf9", "Cd53",
"Xdh", "Hmgb2", "Cxcl14", "Hmox1", "Trim67", "Ppargc1a", "Nlrc5",
"Tnfrsf1a", "Zfpm2", "Eif2ak2", "Birc5", "Eps8", "Stat1", "Ifi27",
"Trim5"), entrezgene = c(15957L, 15945L, 58185L, 15959L, 69550L,
102642251L, 54123L, 13040L, 78781L, 13653L, 100038882L, 14469L,
71586L, 211550L, 12260L, 17289L, 66141L, 13649L, 16362L, 383619L,
12702L, NA, 22339L, 16391L, 12508L, 22436L, 97165L, 57266L, 15368L,
330863L, 19017L, 434341L, 21937L, 22762L, 19106L, 11799L, 13860L,
20846L, 52668L, 667823L), description = structure(c(19L, 6L,
30L, 20L, 3L, 10L, 24L, 4L, 39L, 11L, 26L, 15L, 22L, 33L, 9L,
8L, 21L, 13L, 23L, 1L, 32L, 27L, 37L, 25L, 5L, 38L, 17L, 7L,
16L, 35L, 29L, 28L, 36L, 40L, 14L, 2L, 12L, 31L, 18L, 34L), .Label = c("absent in melanoma 2 [Source:MGI Symbol;Acc:MGI:2686159]",
"baculoviral IAP repeat-containing 5 [Source:MGI Symbol;Acc:MGI:1203517]",
"bone marrow stromal cell antigen 2 [Source:MGI Symbol;Acc:MGI:1916800]",
"cathepsin S [Source:MGI Symbol;Acc:MGI:107341]", "CD53 antigen [Source:MGI Symbol;Acc:MGI:88341]",
"chemokine (C-X-C motif) ligand 10 [Source:MGI Symbol;Acc:MGI:1352450]",
"chemokine (C-X-C motif) ligand 14 [Source:MGI Symbol;Acc:MGI:1888514]",
"c-mer proto-oncogene tyrosine kinase [Source:MGI Symbol;Acc:MGI:96965]",
"complement component 1, q subcomponent, beta polypeptide [Source:MGI Symbol;Acc:MGI:88224]",
"complement component 3 [Source:MGI Symbol;Acc:MGI:88227]", "early growth response 1 [Source:MGI Symbol;Acc:MGI:95295]",
"epidermal growth factor receptor pathway substrate 8 [Source:MGI Symbol;Acc:MGI:104684]",
"epidermal growth factor receptor [Source:MGI Symbol;Acc:MGI:95294]",
"eukaryotic translation initiation factor 2-alpha kinase 2 [Source:MGI Symbol;Acc:MGI:1353449]",
"guanylate binding protein 2 [Source:MGI Symbol;Acc:MGI:102772]",
"heme oxygenase (decycling) 1 [Source:MGI Symbol;Acc:MGI:96163]",
"high mobility group box 2 [Source:MGI Symbol;Acc:MGI:96157]",
"interferon, alpha-inducible protein 27 [Source:MGI Symbol;Acc:MGI:1277180]",
"interferon-induced protein with tetratricopeptide repeats 1 [Source:MGI Symbol;Acc:MGI:99450]",
"interferon-induced protein with tetratricopeptide repeats 3 [Source:MGI Symbol;Acc:MGI:1101055]",
"interferon induced transmembrane protein 3 [Source:MGI Symbol;Acc:MGI:1913391]",
"interferon induced with helicase C domain 1 [Source:MGI Symbol;Acc:MGI:1918836]",
"interferon regulatory factor 1 [Source:MGI Symbol;Acc:MGI:96590]",
"interferon regulatory factor 7 [Source:MGI Symbol;Acc:MGI:1859212]",
"interferon regulatory factor 9 [Source:MGI Symbol;Acc:MGI:107587]",
"ISG15 ubiquitin-like modifier [Source:MGI Symbol;Acc:MGI:1855694]",
"myxovirus (influenza virus) resistance 1 [Source:MGI Symbol;Acc:MGI:97243]",
"NLR family, CARD domain containing 5 [Source:MGI Symbol;Acc:MGI:3612191]",
"peroxisome proliferative activated receptor, gamma, coactivator 1 alpha [Source:MGI Symbol;Acc:MGI:1342774]",
"radical S-adenosyl methionine domain containing 2 [Source:MGI Symbol;Acc:MGI:1929628]",
"signal transducer and activator of transcription 1 [Source:MGI Symbol;Acc:MGI:103063]",
"suppressor of cytokine signaling 3 [Source:MGI Symbol;Acc:MGI:1201791]",
"TRAF-interacting protein with forkhead-associated domain [Source:MGI Symbol;Acc:MGI:2182965]",
"tripartite motif-containing 5 [Source:MGI Symbol;Acc:MGI:3646853]",
"tripartite motif-containing 67 [Source:MGI Symbol;Acc:MGI:3045323]",
"tumor necrosis factor receptor superfamily, member 1a [Source:MGI Symbol;Acc:MGI:1314884]",
"vascular endothelial growth factor A [Source:MGI Symbol;Acc:MGI:103178]",
"xanthine dehydrogenase [Source:MGI Symbol;Acc:MGI:98973]", "zinc finger CCCH type, antiviral 1 [Source:MGI Symbol;Acc:MGI:1926031]",
"zinc finger protein, multitype 2 [Source:MGI Symbol;Acc:MGI:1334444]"
), class = "factor"), gene_biotype = structure(c(1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L), .Label = "protein_coding", class = "factor")), row.names = c(NA,
-40L), .Names = c("ensembl_gene_id", "baseMean", "log2FoldChange",
"lfcSE", "stat", "pvalue", "padj", "gene_symbol", "entrezgene",
"description", "gene_biotype"), class = "data.frame")innateDB_bulk_up_genes <- innateDB_genes_NSCs_bulk_up_table %>% dplyr::arrange( desc( log2FoldChange )) %>% ungroup() %>% dplyr::select(ensembl_gene_id , gene_symbol )
innateDB_bulk_up_genes <- innateDB_bulk_up_genes[1:7,]We split the df into a list , rows become fields
genes_up_InnateDB_sms2_bulk_list <- split(innateDB_bulk_up_genes , f = seq(nrow(innateDB_bulk_up_genes)))Next up we calculate the totalReads and reads for only the specified gene, add the information for cells which age and celltype they belong to.
genes_up_InnateDB_data_list <- lapply(genes_up_InnateDB_sms2_bulk_list, function(x){
raw_count_data <- data.frame( totalReads = colSums(data_10X_NSCs[,as.character(cells_NSCs_10X)]) , geneReads = as.numeric(data_10X_NSCs[ as.character(x$ensembl_gene_id) ,as.character(cells_NSCs_10X)]) , gene = as.character(x$gene_symbol) ) %>% rownames_to_column("name") %>% left_join(anno_10X_NSCs , by = "name" )
table_cells_expr <- addmargins( table( raw_count_data$age , raw_count_data$geneReads > 0 ) )
# fig.width=6 , fig.height=2
gg <- ggplot(data = raw_count_data , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") ) + ggtitle(x$gene_symbol)
list( "raw_count_data" = raw_count_data , "numbers_cells_expressing" = table_cells_expr , "plot" = gg )
})We extract the raw_counts data table from the list we have just created
genes_up_InnateDB_data_list_raw_counts <- lapply(genes_up_InnateDB_data_list, function(x){ x$raw_count_data })and bind together all the data frames for the different genes
genes_up_InnateDB_data_plot_10X <- bind_rows(genes_up_InnateDB_data_list_raw_counts)Scatterplot
gg_InnateDB_10X <- ggplot(data = genes_up_InnateDB_data_plot_10X , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") ) + facet_grid( gene ~ . )gg_InnateDB_10X
Barplot
Show the same data as stacked barplots
genes_up_InnateDB_data_plot_10X_bars <- genes_up_InnateDB_data_plot_10X %>% mutate( interval = base::cut( geneReads , breaks = c(-0.1, 0.5,5.5,50.5,Inf) , labels = c("0","1-5","5-50","50-") ) ) %>% group_by(interval,gene,age) %>% summarise( counts = n() ) %>% ungroup() %>% mutate( interval = fct_rev( interval ) , age = fct_recode( .f = age , "22 month" = "old", "2 month" = "young" ) %>% fct_rev() ) %>% group_by(gene,age) %>% mutate( Percent = 100*counts/sum(counts) )ggplot( data = genes_up_InnateDB_data_plot_10X_bars , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ gene ) + viridis::scale_fill_viridis( discrete = TRUE , direction = -1 , drop=FALSE , name = "Read count" ) + ggtitle("Read counts: 10X") + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5) , strip.text.x = element_text(size = 8)) 
SMARTseq2
How many counts do genes have in the Single cell data sets, that are significantly upregulated in the bulk NSCs
cells_NSCs_SMARTseq2 <- as.character(anno_SMARTseq2_NSCs$name)Ifit1 - ENSMUSG00000078920
Ifit1_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000078920",cells_NSCs_SMARTseq2]) , gene = "Ifit1" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ifit1_df_SMARTseq2$age , Ifit1_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 132 1 133
young 89 0 89
Sum 221 1 222gg_Ifit1_SMARTseq2 <- ggplot(data = Ifit1_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ifit1_SMARTseq2
Rsad2 - ENSMUSG00000020641
Rsad2_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000020641",cells_NSCs_SMARTseq2]) , gene = "Rsad2" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Rsad2_df_SMARTseq2$age , Rsad2_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 122 11 133
young 86 3 89
Sum 208 14 222gg_Rsad2_SMARTseq2 <- ggplot(data = Rsad2_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Rsad2_SMARTseq2
C3 - ENSMUSG00000024164
C3_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000024164",cells_NSCs_SMARTseq2]) , gene = "C3" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( C3_df_SMARTseq2$age , C3_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 133 0 133
young 88 1 89
Sum 221 1 222gg_C3_SMARTseq2 <- ggplot(data = C3_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_C3_SMARTseq2
Iigp1 - ENSMUSG00000054072
Iigp1_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000054072",cells_NSCs_SMARTseq2]) , gene = "Iigp1" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Iigp1_df_SMARTseq2$age , Iigp1_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 116 17 133
young 89 0 89
Sum 205 17 222gg_Iigp1_SMARTseq2 <- ggplot(data = Iigp1_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Iigp1_SMARTseq2
Ifi44 - ENSMUSG00000028037
Ifi44_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000028037",cells_NSCs_SMARTseq2]) , gene = "Ifi44" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ifi44_df_SMARTseq2$age , Ifi44_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 126 7 133
young 88 1 89
Sum 214 8 222gg_Ifi44_SMARTseq2 <- ggplot(data = Ifi44_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ifi44_SMARTseq2
Cxcl10 - ENSMUSG00000034855
Cxcl10_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000034855",cells_NSCs_SMARTseq2]) , gene = "Cxcl10" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Cxcl10_df_SMARTseq2$age , Cxcl10_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 130 3 133
young 89 0 89
Sum 219 3 222gg_Cxcl10_SMARTseq2 <- ggplot(data = Cxcl10_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Cxcl10_SMARTseq2
Il33 - ENSMUSG00000024810
Il33_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000024810",cells_NSCs_SMARTseq2]) , gene = "Il33" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Il33_df_SMARTseq2$age , Il33_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 58 75 133
young 54 35 89
Sum 112 110 222gg_Il33_SMARTseq2 <- ggplot(data = Il33_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Il33_SMARTseq2
Ddx60 - ENSMUSG00000037921
Ddx60_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000037921",cells_NSCs_SMARTseq2]) , gene = "Ddx60" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ddx60_df_SMARTseq2$age , Ddx60_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 113 20 133
young 87 2 89
Sum 200 22 222gg_Ddx60_SMARTseq2 <- ggplot(data = Ddx60_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ddx60_SMARTseq2
Ifi47 - ENSMUSG00000078920
Ifi47_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000078920",cells_NSCs_SMARTseq2]) , gene = "Ifi47" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Ifi47_df_SMARTseq2$age , Ifi47_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 132 1 133
young 89 0 89
Sum 221 1 222gg_Ifi47_SMARTseq2 <- ggplot(data = Ifi47_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Ifi47_SMARTseq2
Gm4951 - ENSMUSG00000073555
Gm4951_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000073555",cells_NSCs_SMARTseq2]) , gene = "Gm4951" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Gm4951_df_SMARTseq2$age , Gm4951_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 120 13 133
young 89 0 89
Sum 209 13 222gg_Gm4951_SMARTseq2 <- ggplot(data = Gm4951_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Gm4951_SMARTseq2
Slc1a3 - ENSMUSG00000005360
Slc1a3_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000005360",cells_NSCs_SMARTseq2]) , gene = "Slc1a3" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Slc1a3_df_SMARTseq2$age , Slc1a3_df_SMARTseq2$geneReads > 0 ) )
TRUE Sum
old 133 133
young 89 89
Sum 222 222gg_Slc1a3_SMARTseq2 <- ggplot(data = Slc1a3_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Slc1a3_SMARTseq2
Sox2 - ENSMUSG00000074637
Sox2_df_SMARTseq2 <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs) , geneReads = as.numeric(data_SMARTseq2_NSCs["ENSMUSG00000074637",cells_NSCs_SMARTseq2]) , gene = "Sox2" ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )Column `name` joining character vector and factor, coercing into character vectoraddmargins( table( Sox2_df_SMARTseq2$age , Sox2_df_SMARTseq2$geneReads > 0 ) )
FALSE TRUE Sum
old 48 85 133
young 25 64 89
Sum 73 149 222gg_Sox2_SMARTseq2 <- ggplot(data = Sox2_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") )
gg_Sox2_SMARTseq2
Combine all genes
interferon_genes_bulk_df_SMARTseq2 <- bind_rows( Ifit1_df_SMARTseq2 , C3_df_SMARTseq2 , Rsad2_df_SMARTseq2 , Iigp1_df_SMARTseq2 , Ifi47_df_SMARTseq2 , Il33_df_SMARTseq2 , Cxcl10_df_SMARTseq2 , Sox2_df_SMARTseq2 , Slc1a3_df_SMARTseq2 )Unequal factor levels: coercing to characterbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorbinding character and factor vector, coercing into character vectorScatterplot
gg_interferon_genes_SMARTseq2 <- ggplot(data = interferon_genes_bulk_df_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") ) + facet_grid( gene ~ . ) + scale_y_log10() + ylab("log10( geneReads )")gg_interferon_genes_SMARTseq2
Barplot
Show the same data as stacked barplots
interferon_genes_bulk_df_SMARTseq2_bars <- interferon_genes_bulk_df_SMARTseq2 %>% mutate( interval = base::cut( geneReads , breaks = c(-0.1, 0.5,5.5,50.5,Inf) , labels = c("0","1-5","5-50","50-") ) ) %>% group_by(interval,gene,age) %>% summarise( counts = n() ) %>% ungroup() %>% mutate( interval = fct_rev( interval ) , age = fct_recode( .f = age , "22 month" = "old", "2 month" = "young" ) %>% fct_rev() ) %>% group_by(gene,age) %>% mutate( Percent = 100*counts/sum(counts) )ggplot( data = interferon_genes_bulk_df_SMARTseq2_bars , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ gene ) + viridis::scale_fill_viridis( discrete = TRUE , direction = -1 ) + ggtitle("Read counts: SMARTseq2")
Bulk NSC genes in InnateDB
Which genes are the ones with high fold changes in the bulk and are annotated in InnateDB
innateDB_bulk_up_genes <- innateDB_genes_NSCs_bulk_up_table %>% dplyr::arrange( desc( log2FoldChange )) %>% ungroup() %>% dplyr::select(ensembl_gene_id , gene_symbol )
innateDB_bulk_up_genes <- innateDB_bulk_up_genes[1:7,]We split the df into a list , rows become fields
genes_up_InnateDB_sms2_bulk_list <- split(innateDB_bulk_up_genes , f = seq(nrow(innateDB_bulk_up_genes)))Next up we calculate the totalReads and reads for only the specified gene, add the information for cells which age and celltype they belong to.
genes_up_InnateDB_data_list_SMARTseq2 <- lapply(genes_up_InnateDB_sms2_bulk_list, function(x){
raw_count_data <- data.frame( totalReads = colSums(data_SMARTseq2_NSCs[,as.character(cells_NSCs_SMARTseq2)]) , geneReads = as.numeric(data_SMARTseq2_NSCs[ as.character(x$ensembl_gene_id) ,as.character(cells_NSCs_SMARTseq2) ]) , gene = as.character( x$gene_symbol ) ) %>% rownames_to_column("name") %>% left_join(anno_SMARTseq2_NSCs , by = "name" )
table_cells_expr <- addmargins( table( raw_count_data$age , raw_count_data$geneReads > 0 ) )
# fig.width=6 , fig.height=2
gg <- ggplot(data = raw_count_data , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") ) + ggtitle(x$gene_symbol)
list( "raw_count_data" = raw_count_data , "numbers_cells_expressing" = table_cells_expr , "plot" = gg )
})We extract the raw_counts data table from the list we have just created
genes_up_InnateDB_data_list_SMARTseq2_raw_counts <- lapply(genes_up_InnateDB_data_list_SMARTseq2, function(x){ x$raw_count_data })and bind together all the data frames for the different genes
genes_up_InnateDB_data_plot_SMARTseq2 <- bind_rows(genes_up_InnateDB_data_list_SMARTseq2_raw_counts)Scatterplot
gg_InnateDB_SMARTseq2 <- ggplot(data = genes_up_InnateDB_data_plot_SMARTseq2 , mapping = aes(x = totalReads , y = geneReads ) ) + geom_point(aes(color = age)) + scale_color_manual(values = c("old" = "slateblue" , "young" = "yellowgreen") ) + facet_grid( gene ~ . )gg_InnateDB_SMARTseq2
Barplot
Show the same data as stacked barplots
genes_up_InnateDB_data_plot_SMARTseq2_bars <- genes_up_InnateDB_data_plot_SMARTseq2 %>% mutate( interval = base::cut( geneReads , breaks = c(-0.1, 0.5,5.5,50.5,Inf) , labels = c("0","1-5","5-50","50-") ) ) %>% group_by(interval,gene,age) %>% summarise( counts = n() ) %>% ungroup() %>% mutate( interval = fct_rev( interval ) , age = fct_recode( .f = age , "22 month" = "old", "2 month" = "young" ) %>% fct_rev() ) %>% group_by(gene,age) %>% mutate( Percent = 100*counts/sum(counts) )ggplot( data = genes_up_InnateDB_data_plot_SMARTseq2_bars , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ gene ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle("Read counts: SMARTseq2") + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
Combined Barplots for 10X and SMARTseq2
genes_up_InnateDB_data_barplot <- bind_rows(
dplyr::mutate( genes_up_InnateDB_data_plot_SMARTseq2_bars , seqmethod = "Smart-Seq2" ) ,
dplyr::mutate( genes_up_InnateDB_data_plot_10X_bars , seqmethod = "10X" )
)idx = 1
gg1 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[[idx]] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg1
idx = 2
gg2 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[idx] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg2
idx = 3
gg3 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[idx] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg3
idx = 4
gg4 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[idx] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg4
idx = 5
gg5 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[idx] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg5
idx = 6
gg6 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[idx] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg6
idx = 7
gg7 <- ggplot( data = genes_up_InnateDB_data_barplot %>% dplyr::filter( gene == innateDB_bulk_up_genes$gene_symbol[idx] ) , mapping = aes(x = age , y = Percent , fill = interval) ) + geom_bar(stat = "identity") + facet_grid(facets = . ~ seqmethod ) + viridis::scale_fill_viridis( discrete = TRUE , drop=FALSE , direction = -1 , name = "Read count" ) + ggtitle( paste("Read counts:", innateDB_bulk_up_genes$gene_symbol[[idx]] ) ) + theme( axis.text.x = element_text(angle = 90,hjust = 0,vjust = 0.5))
gg7
gridExtra::grid.arrange(gg1, gg2, gg3 , gg4, gg5, gg6 , gg7 , nrow = 1)
SessionInfo
sessionInfo()R version 3.4.3 (2017-11-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 16.04.3 LTS
Matrix products: default
BLAS: /usr/lib/libblas/libblas.so.3.6.0
LAPACK: /usr/lib/lapack/liblapack.so.3.6.0
locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8 LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] parallel stats graphics grDevices utils datasets methods base
other attached packages:
[1] biomaRt_2.34.2 ggbeeswarm_0.6.0 viridisLite_0.3.0 bindrcpp_0.2 forcats_0.3.0 dplyr_0.7.4
[7] ggrepel_0.8.0 Seurat_2.2.0 Biobase_2.38.0 BiocGenerics_0.24.0 Matrix_1.2-14 cowplot_0.7.0
[13] ggplot2_2.2.1 pheatmap_1.0.10 tibble_1.4.2 stringr_1.3.1 tidyr_0.7.2
loaded via a namespace (and not attached):
[1] readxl_1.1.0 backports_1.1.2 Hmisc_4.1-1 VGAM_1.0-3 NMF_0.20.6
[6] sn_1.5-0 plyr_1.8.4 igraph_1.0.1 lazyeval_0.2.0 splines_3.4.3
[11] gridBase_0.4-7 digest_0.6.12 foreach_1.4.3 htmltools_0.3.6 viridis_0.4.1
[16] lars_1.2 gdata_2.17.0 memoise_1.0.0 magrittr_1.5 checkmate_1.8.5
[21] cluster_2.0.6 doParallel_1.0.10 mixtools_1.0.4 ROCR_1.0-7 openxlsx_4.1.0
[26] R.utils_2.6.0 prettyunits_1.0.2 colorspace_1.3-2 blob_1.1.1 haven_1.1.2
[31] RCurl_1.95-4.10 crayon_1.3.4 jsonlite_1.5 lme4_1.1-18-1 bindr_0.1
[36] survival_2.41-3 iterators_1.0.8 ape_5.1 glue_1.3.0 registry_0.3
[41] gtable_0.2.0 MatrixModels_0.4-1 car_3.0-2 kernlab_0.9-25 prabclus_2.2-6
[46] DEoptimR_1.0-8 abind_1.4-5 scales_0.5.0 mvtnorm_1.0-5 DBI_1.0.0
[51] rngtools_1.2.4 Rcpp_0.12.18 dtw_1.20-1 progress_1.2.0 xtable_1.8-2
[56] htmlTable_1.12 tclust_1.2-3 bit_1.1-14 foreign_0.8-69 proxy_0.4-22
[61] mclust_5.2.2 SDMTools_1.1-221 Formula_1.2-3 tsne_0.1-3 stats4_3.4.3
[66] httr_1.3.1 htmlwidgets_1.2 FNN_1.1 gplots_3.0.1 RColorBrewer_1.1-2
[71] fpc_2.1-10 acepack_1.4.1 modeltools_0.2-22 ica_1.0-2 XML_3.98-1.11
[76] pkgconfig_2.0.2 R.methodsS3_1.7.1 flexmix_2.3-13 nnet_7.3-12 caret_6.0-73
[81] AnnotationDbi_1.40.0 labeling_0.3 rlang_0.2.2 reshape2_1.4.2 munsell_0.4.3
[86] cellranger_1.1.0 tools_3.4.3 RSQLite_2.1.1 ranger_0.6.0 ggridges_0.5.0
[91] evaluate_0.11 yaml_2.2.0 bit64_0.9-7 ModelMetrics_1.1.0 knitr_1.20
[96] zip_1.0.0 robustbase_0.92-7 caTools_1.17.1 purrr_0.2.4 pbapply_1.3-1
[101] nlme_3.1-131 R.oo_1.22.0 compiler_3.4.3 rstudioapi_0.7 beeswarm_0.2.3
[106] curl_3.2 stringi_1.2.4 lattice_0.20-35 trimcluster_0.1-2 nloptr_1.0.4
[111] diffusionMap_1.1-0.1 pillar_1.3.0 data.table_1.11.6 bitops_1.0-6 irlba_2.1.2
[116] R6_2.2.2 latticeExtra_0.6-28 KernSmooth_2.23-15 gridExtra_2.2.1 rio_0.5.10
[121] IRanges_2.12.0 vipor_0.4.5 codetools_0.2-15 boot_1.3-20 MASS_7.3-48
[126] gtools_3.5.0 assertthat_0.2.0 pkgmaker_0.22 rprojroot_1.3-2 mnormt_1.5-5
[131] S4Vectors_0.16.0 diptest_0.75-7 hms_0.4.2 grid_3.4.3 rpart_4.1-12
[136] minqa_1.2.4 class_7.3-14 rmarkdown_1.10 segmented_0.5-1.4 carData_3.0-1
[141] Rtsne_0.11 numDeriv_2016.8-1 scatterplot3d_0.3-41 base64enc_0.1-3 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